Are patients' decision‐making preferences being met?

Abstract

Objective  To investigate the information and decision‐making expectations of general practice patients during real life consultations.

Design  Post‐consultation, quantitative patient preference and enablement questionnaire.

Setting and participants  Patients attending for routine appointments in general practice surgeries in Oxfordshire, UK.

Results  Thirteen Oxfordshire general practitioners (GPs) volunteered to take part and a total of 171 patients completed and returned the questionnaire. Between a quarter and one‐third of patients reported receiving less information than they desired, particularly in relation to the risks and benefits of medical treatments. Patients who preferred the doctor to make decisions for them (35%), were more likely to have their preferences met (64%) compared with patients wishing to share decisions (47%) or make their own (18%) who were less likely to achieve this role (52 and 41%, respectively). However, it could not be demonstrated unequivocally that these differences were statistically significant. In total, 61% of patients perceived that they achieved their preferred decision‐making role. No significant differences were found in post‐consultation enablement scores between any of the decision preference groups. Patients' assessments indicated that some doctors were more successful at achieving congruence than others.

Conclusion  The decision‐making preferences of general practice patients tend to vary. However, there was a substantial mismatch between the stated preferences of patients for the role they wanted to have in decision‐making and what they felt actually took place in their consultation. Therefore, it remains a challenge for doctors to match their consultation style to the decision‐making preferences of individual patients.

Introduction

In western medical practice, patient involvement in health‐care is assuming greater and greater importance. Increasingly, there is a demand for patients to be provided with information about their medical condition and available treatment options to enable them to make informed choices and participate in decisions about their care. In the UK, National Health Service (NHS) initiatives advocate ‘active partnerships’ between health professionals and patients, and the improvement of information based on good quality evidence to help patients choose between options. ¹ , ² The NHS Plan emphasizes the need for patients to have more say in their own care and more influence over the way in which the NHS works. ³ The move towards increasing patient involvement is driven by a theoretical concern for respect for patient autonomy. This emphasizes that patients should be in a position to choose whether to accept an intervention or not as part of their general right to determine their own lives. ⁴

The growing body of international research suggests that giving health information to patients and involving them in decisions about their medical care can result in beneficial psychological and physical outcomes. ⁵ However, the outcomes for patients who take part in decision‐making are not always positive. ⁶ , ⁷ Furthermore, some professionals believe that most patients do not want to make decisions about their medical care and would prefer to leave it up to the doctor. ⁸ , ⁹ There is less evidence about how primary‐care physicians respond to the preferences of patients for information and involvement.

Studies that investigate whether patients actually wish to participate in decision‐making tend to survey different populations using different measures. ¹⁰ Some focus on hospital patients facing life‐threatening illness such as cancer, ¹¹ others question healthy members of the public, ¹² and many use hypothetical situations. ¹³ The current evidence indicates that patients' desire for information is stronger than their desire to be involved in decision‐making. ¹⁰ It seems that older patients, tend to favour a more directive doctor style, whereas younger and better‐educated patients prefer more involvement in decision‐making. ¹⁴ There is also an indication that patients are more inclined to become involved in decision‐making when they have sufficient information about their treatment options and the likely consequences of these. ¹⁰ , ¹⁵

These general conclusions are based on research conducted predominantly in secondary‐care settings. Although there is evidence that general practitioners (GPs) consistently underestimate patients' desire for information and seldom elicit treatment preferences, ¹⁶ , ¹⁷ comparatively little is known about the decision‐making preferences of primary‐care patients. The few studies conducted have yielded conflicting results. For example, Little et al. ¹⁸ found that most patients prefer a patient‐centred approach to care. In their study of 865 general practice patients, they identified three domains of patient preferences agreed by at least 85% of patients. These were communication, partnership and health promotion. ¹⁸ In a US study, Bradley et al. ¹⁹ used a questionnaire comprising one general question and 10 different clinical scenarios, followed by a list of seven options that described varying degrees of patient participation in decision‐making. Patient responses to seven of the scenarios were statistically different from their responses predicted by the general preference question. Conversely, the authors found that patients generally preferred doctors to play the dominant role in decision‐making. In Scotland, McKinstry ²⁰ used structured interviews with video vignettes of acted consultations to determine the decision‐making preferences of 410 general practice patients. Pairs of vignettes of five common scenarios were made, one using a shared and the other a directed approach. Similarly, the study showed that a large number of patients preferred directed consultations when viewing the scenarios. The results of another vignette study investigating Australian, UK and US preferences for participation in medical decision‐making suggested that mutuality was the preferred method of decision‐making for participants in all three countries. ²¹ Respondents in the UK and US preferred delegating decisions rather than deciding alone, while Australian participants preferred to decide alone rather than to delegate. The UK and US samples, however, were drawn from primary‐care settings, whereas the Australian participants were undergraduate students from health faculties at the University of Western Sydney.

Most of these studies (excluding the study by Little et al.) involved asking patients about hypothetical situations. The data reported here are derived from general practice patients attending for real consultations. There were three main aims. First, to assess patient expectations of the consultation in relation to receipt of medical information and decision‐making role. Secondly, to obtain patient assessments of whether these expectations were realized. Finally, to assess whether the fulfillment of patient expectations influences patients' feelings of enablement after a general practice consultation.

Participants and methods

The research protocol was approved by the Oxfordshire Applied and Qualitative Research Ethics Committee (AQREC). Information about the study and invitations to participate were sent to 40 Oxfordshire GPs. Doctors were informed that the purpose of the study was to investigate opportunities for patient involvement in medical decision‐making. Those who had recently attended educational workshops at the Institute of Health Sciences were specifically targeted as it was likely that they would be interested in general practice research. A total of 13 doctors volunteered to participate from 12 different practices. Four practices were in deprived city areas, four in rural areas and the remaining four in market towns with patients from urban and rural areas. Consecutive patients attending each surgery were invited to participate. On entering the practice waiting room, they were given a written information sheet by the first author. This sheet explained that some patients felt that they would like to be more involved in the decisions made about their treatment and care and that the study was investigating ways in which doctors could give good‐quality, up‐to‐date information to their patients. After being given the opportunity to ask questions about the study, those willing to participate were asked for their written consent. Although no formal record of refusals was made, we estimate that 10–15% of patients who were approached preferred not to participate. Participation involved completing a post‐consultation decision preference questionnaire. All patients capable of completing the questionnaire were eligible to participate. Sociodemographic details and data concerning the number and type of agenda items raised during the consultation were also collected.

Questionnaire

After their consultation, participants completed a questionnaire that examined their expectations and perceptions concerning their visit to the doctor. If patients were unable to complete this before they left the surgery, they were given a stamped addressed envelope in which to return the form after completion at a more convenient time. The questionnaire was derived from three existing instruments used previously in communication research to measure: patient enablement, ²² desired decision‐making role, ²³ and information preferences and perceptions of the actual consultation. ¹⁶ In section 1 patients were asked to agree or disagree on a six‐point continuous scale (‘very strongly agree’ to ‘very strongly disagree’) with statements about how much information they wanted concerning their condition and treatment options. Section 2 asked them to describe their general preference concerning making decisions about their medical care and treatment. The questionnaire stated that treatment included prescribed medication, self‐care and referral to hospital. Patients were asked to tick one preferred option from a choice of five which were worded as follows:

‘I prefer to make the decision(s) about which treatment I will receive’.

‘I prefer to make the decision(s) about my treatment after seriously considering my doctor's opinion'.

‘I prefer that my doctor and I share responsibility for deciding which treatment is best for me’.

‘I prefer that my doctor makes the final decision about which treatment I will receive, but seriously considers my opinion’.

‘I prefer to leave all decisions regarding treatment to my doctor’. ²³

In section 3 patients were asked to complete the enablement scale. This validated scale contains six general questions about patients' experience of the consultation. Enablement is complementary to, but different from, satisfaction and places importance on concepts such as patient‐centredness, empowerment and on patients' ability to understand and cope with illness. ²² Section 4 asked participants to state how much information they felt they had actually received during their consultation concerning their condition and treatment (also rated on a continuous scale as in section 1). Section 5 asked patients who they thought made the decisions during their consultation, i.e. was it the doctor, were decisions shared, or did the patient make his/her own decision. Patients were asked to tick one of the six choices (there was an extra choice of ‘no decision was made’).

We chose to administer the patient preference questionnaire after the consultation, as we did not want to raise patient consciousness in any way. If patients were asked what level of input they expected in the decision‐making process, before consultation, it is possible that this would have influenced their behaviour during the consultation and we wanted to avoid this. Our view is that this problem of influence is more important than the value in detecting change.

Data analysis

The patient self‐report data were entered into SPSS for Windows (version 10). Statistical analyses included descriptive procedures and one‐way analyses of variance tests to investigate differences between patient enablement scores and different decision‐making preference groups. A chi‐square test was conducted to compare the differences in number of patients in each decision preference group (i.e. stated preference and perception of actual role played). The Kappa statistic option for this test was selected to assess the level of agreement between patients' desired decision‐making role, and what they perceived they actually received. Patients who had at least one new problem requiring treatment or at least one established problem requiring either a new or change in treatment were included in the main analyses.

Results

A total of 212 patients agreed to take part in the study and of these, the total number who returned completed questionnaires was 171 (78%). Patients who declined to take part were often short of time and in a hurry to leave the surgery. There were 19 patients who attended for a review only of a long‐term problem and these were excluded from the main analyses, as decisions regarding treatment did not arise in these encounters.

Table 1 lists the characteristics of the sample. This comprised 67% female patients and the mean age was 49 (median 46, range 16–88). The majority of agenda items discussed were of a physical nature with 61% of patients attending for at least one new problem. On average each GP treated 12 patients who were participating in the study.

Table 1.

Sample and consultation characteristics (n = 171)

	n (%)
Level of education
Some secondary school	10 (5.4)
Secondary school	88 (51.6)
Technical college	28 (16.3)
University	30 (17.9)
Postgraduate qualification	15 (8.7)
Employment status
Working full‐time	60 (35.4)
Retired	51 (30.2)
Unemployed	30 (17.4)
Working part‐time	25 (14.3)
Student	5 (2.6)
Consultation agenda items
Physical	138 (80.4)
Psychological/emotional	12 (7.1)
Health promotion	11 (6.1)
Administrative	6 (3.8)
Social/domestic	4 (2.6)
New: requiring treatment	104 (61.1)*
Established: new/Δ treatment	66 (38.9)*
Established: review only	35 (21.0)*

^* At least one agenda item.

Health and treatment information preferences

Eighty‐seven per cent of all patients with at least one new problem needing treatment strongly agreed that they wanted the doctor to give them all the information they needed about ways of managing their problems, whilst 74% stated that they had actually received this information during the consultation.

Receipt of information about the benefits and risks of proposed treatments was important to patients. Eighty‐four per cent strongly agreed that they wanted the doctor to explain how the treatment would benefit them and 83% preferred to receive information about the risks of the treatment. However, 70 and 62% of these patients, respectively, felt that they had actually received this information. Receiving scientific evidence regarding the effectiveness of treatment was seen as less important with 41% strongly agreeing that they wanted their doctor to provide them with this information and 67% of these patients stating that they had received this. Therefore, patients reported receiving less information than they wanted concerning their condition and treatment.

General decision‐making preferences

To strengthen the analysis, the decision‐making preference response items 1–2 and 4–5 were collapsed into two separate categories making a total of three categories. These are presented in Table 2, together with the enablement scores for each decision preference group.

Table 2.

Decision‐making preferences, self‐reported outcome and enablement scores: patients with established and new problems needing treatment

General patient preferences	n = 152 n (%)†	Doctor made decision		PEI mean score	Decision was shared		PEI mean score	I made decision		PEI mean score	No decision made		PEI mean score
		n	%		n	%		n	%		n	%
I prefer decisions to be made by the doctor	53 (35)	34	64.1	4.26	12	22.6	1.75	4	7.5	4.26	3	5.8	4.33
I prefer to share decisions with the doctor	72 (47)	20	27.8	6.05	38	52.8	4.26	5	6.9	6.20	9	12.5	2.22
I prefer to make my own decisions	27 (18)	10	37.0	4.90	4	14.8	5.25	11	40.8	5.00	2	7.4	2.00

^† Figures in brackets represent the number of patients in each of the three preference groups as a percentage of the total sample.

All other percentages are based on the total n for each group (not the sample as a whole).

Just over one‐third (35%) of patients included in the analysis stated that, in general, they preferred the doctor to make the decisions about their treatment, with the remaining two‐thirds preferring either to share (47%) or make their own decisions (18%). Those patients who preferred the doctor to make decisions for them were significantly older than those who preferred to make their own decisions (51.8 vs. 39.8, P = 0.04).

In 47% of all consultations, patients reported that decisions were made by the doctor alone. In 39%, shared decision‐making took place and in 14% of consultations, patients perceived that they made their own decisions. Of those patients who stated a preference for the doctor to make the decision for them; 64.1% reported that the doctor actually made the decision for them. 52.8% of patients who preferred a shared approach agreed that they had shared the decision with the doctor and 40.8% of patients wanting to make their own decisions felt that they had made the decision themselves. A chi‐square test suggested that these were not chance differences [χ ² = 40.0 (4 d.f.), P < 0.001]. However, as one cell had an expected frequency of <5, this result must be interpreted with extreme caution. In 9% of cases overall, it was stated that no decision had been made. On aggregate, the number of patients who felt that they achieved their desired decision role was 61%. The Kappa coefficient for the level of agreement between patients' decision‐making expectations and their assessments of what actually happened was 0.36. Patients' assessments of the decision‐making process during the consultation suggest that some doctors were more successful at meeting their patients' expectations than others. The average number of patients per doctor who perceived that their decision‐making preferences had been met was 55%– this ranged from 23 to 92%.

Enablement

The average enablement score across all patients was 4.4 (range 0–12). Table 2 shows the mean scores for each decision preference group. Those patients who preferred to make their own decisions and felt they achieved this had a higher mean score (+0.74) than both the other groups. However, this was not significant. It was not possible to test for significant differences across all groups due to diminutive cell sizes. Consequently, no other significant differences in post‐consultation enablement scores were found between decision preference groups.

Discussion

This study used patient self‐reports to investigate unselected general practice patients' information and decision‐making preferences and to assess whether their expectations were met during real primary‐care consultations. The findings reveal that between a quarter and one‐third of patients did not receive as much health information as they wanted, particularly in relation to the risks and benefits of treatment. This accords with previous research concerning patients' preferences for information.

The main finding of the study is that there was a substantial mismatch between the stated preferences of patients for the role they would like to take in decision‐making and what they felt actually took place in their consultation. While just under two‐thirds of patients stated a preference for either a shared approach or taking sole responsibility, in nearly half of all consultations the patient felt that the decision was made by the doctor alone. Thirty‐nine per cent of patients overall perceived that they did not achieve their preferred decision‐making role, and over half of patients who wanted to make or share decisions felt that they had not been involved.

Patients who preferred to have decisions made for them tended to be older than those who preferred more involvement. This accords with a review of published surveys by Benbassat. ²⁴ It could be that this older age group who are generally used to and perhaps more comfortable with a paternalistic doctor–patient relationship, view shared decision‐making as indicative of ‘doctor uncertainty’ and for this reason feel less satisfied when faced with this approach. ²⁵ Furthermore, if the chance of involvement is unexpected it could have an anxiety‐provoking effect on these patients.

Although the numbers of patients seen by each doctor in the study were small, it appeared that some doctors were more successful at achieving congruity between their patients' expectations and perceived role achievement than others. There is evidence that some GPs and junior clinicians do not possess the skills needed to elicit patients' preferences and involve them in clinical decision‐making. ¹⁵ , ²⁶ It may be that this degree of congruence could be further developed as a future index of quality in the consultation.

Just over half of those patients who wanted to share decision‐making felt that they did actually share decisions during their consultation. Recent research conducted by Stevenson et al. ⁹ suggests reasons for the absence of shared decision‐making in many general practice consultations. These authors found a lack of information sharing between doctors and patients concerning treatments. Patients' beliefs were not taken seriously so that there was little basis upon which to build a consensus and share decisions about which treatments to choose.

There was no firm evidence to suggest that achieving role preference had an enabling effect on patients. This we believe is mostly due to the final sample size which reduced the power of the study, thus making it unlikely to assess anything but large changes in enablement scores. However, we are also aware that the Patient Enablement Instrument may require more psychometric development as an outcome measure for general practice consultations. ²⁶

Limitations of the study

We are aware that the relatively smaller number of patients with psychosocial problems makes our general practice sample untypical. Based on our past experience of research in general practice, it is precisely these types of emotionally vulnerable patients who are less likely to give their consent to take part in this type of study.

Our research is also limited by the fact that it was an investigation of patients' general decision‐making preferences and there is evidence that patient preferences for decision‐making can vary depending on the presenting problem. For example, some patients with mental health‐ and lifestyle‐problems want their views on management to be taken more into account, whilst those with physical problems prefer a more directed approach. ²⁰ Furthermore, this was an investigation of patients' perceptions, rather than an objective assessment of what actually happened during the consultation. Previous research suggests that patients over‐estimate the amount of information they are given during the consultation concerning their treatment options. ¹⁶

Moreover, we have no tangible data from patients concerning what in their view constitutes a shared decision. There are a number of recently formulated theoretical approaches to the consultation which emphasize respect for patient involvement in health‐care decisions and the sharing of good‐quality information. ²⁷ , ²⁸ These have been developed by health professionals and academics some of whom have proposed the principles and competencies for shared decision‐making in the medical encounter. ²⁹ , ³⁰ , ³¹ As yet we do not know whether the consultations of those patients who perceived that they shared decision‐making, bear any resemblance to the models put forward by these authors. It is likely that the patient's view of shared decision‐making is quite different from that of the health professional. We also acknowledge that using a questionnaire‐based approach to investigate patients' preferred level of involvement assumes that they have an understanding of the concepts and questions that they are presented with. Therefore, the use of surveys generally has limitations for gaining an in‐depth understanding of the patient's viewpoint regarding this area of research.

Although this study emphasizes the necessary bridge between prior expectations of patients, actual consultation process and perceived engagement in decision‐making, it does not provide an analysis of actual consultation process. Future ‘input–process–outcome’ studies in this area are needed to fill this gap and provide more insight concerning the pathways between expectations, process, patients' perceived outcomes and objective measures of outcome.

Conclusion

The challenge for clinical practice is how to bridge the gap between patients' preferences for involvement and what they actually receive. In this regard, there is a need for doctors to receive effective training that will equip them with the interpersonal skills needed to combine evidence‐based information sharing with patient‐centred care. This includes the ability to communicate probabilistic information about management options in terms that those patients who wish to participate in decision‐making will understand. Perhaps the first crucial step might be to teach trainees how to elicit the information and involvement preferences of individual patients and how to adapt their consultation style to these preferences so that a good match can be realized for most consultations.

Acknowledgements

We thank all the doctors and patients who volunteered to participate in this study. This paper forms part of a project originally funded by the NHS Anglia and Oxfordshire Regional R&D. We gratefully acknowledge the Dulverton Trust for their continued financial support.